Give your agent a program, not a mouse.

Your agent, or a person, does a task once in an app. OpenAdapt compiles it into a program your agent can call. A healthy run makes no model calls; the program checks the record before and after each save, and stops when a check fails.

Make a brief you can hand to your AI.

Type a job or pick an example to see how OpenAdapt would run it.

Describe the task. Don't enter patient data or personal information, such as names, dates of birth, or phone or record numbers. You get a page with a brief for your AI and the install steps. We keep the job text to see what people ask for.

Try:

A brief is a handoff to your AI. This receipt is from a separate test run.

See a test run's receipt
OpenEMR test copy · Referral 40 model calls
OpenEMR during referral 4, step 28: Click Create New Patient
Stopped before savingOpenEMR listed a possible duplicate where the recording had none. No chart was created.Open its receipt

The program29 steps, compiled from one recording

Name

Date of birth and sex

Address

Phone, email and country

Create the chart

  1. Saved and checked 11 of 11 record checks confirmed
  2. Saved and checked 11 of 11 record checks confirmed
  3. Saved and checked 11 of 11 record checks confirmed
  4. Stopped before saving Possible duplicate patient
Screenshot of each step from four independent runs on a test copy of OpenEMR with synthetic patients, 20 Nov 2026 UTC, sped up. Select a referral or a step to jump to it; select an outcome to open that run’s signed receipt.

See every run, and what needs you

OpenAdapt Cloud shows how each run ended, which runs stopped for a person, a signed receipt for each checked run, and the last two weeks at a glance.

Screenshots of OpenAdapt Cloud with a fictional company and synthetic data.Open the dashboard

The program stopped before saving
a possible duplicate

On referral 4, OpenEMR listed a patient who might be the same person. The program created no chart and left the decision to a person.

OpenEMR after referral 1: the new patient's chart, created by the program
Screenshot · referral 1 · synthetic patientSaved, then read back from OpenEMR11 of 11 fields match the referral.Open its receipt
OpenEMR during referral 4: a possible-duplicate list with one matching patient, and the button to confirm a new chart, which the program did not press
Screenshot · referral 4 · synthetic patientStopped before savingAt “Click Create New Patient”, OpenEMR found a matching chart. No duplicate was created.Open its receipt

The same kind of stop on an earlier run of the referral program, filmed.

Real recordings, sped up. Synthetic patients.

The program it compiled

A recorded referral, compiled into 29 fixed steps. Four runs play through them; the fourth stops at a possible duplicate.

Compiled programNew patient from a referral · OpenEMR test copy
Run trace bound · 4 referrals, recorded
Ready
Compiled topology31 nodes · 31 exact edges · 7 cards
declared loopnot confirmed
  • For each referral in the list to Name (loop body: declared loop)
  • Name to Date of birth and sex (sequence)
  • Date of birth and sex to Address (sequence)
  • Address to Phone, email and country (sequence)
  • Phone, email and country to Create the chart (sequence)
  • Create the chart to For each referral in the list, returns to an earlier step (sequence)
  • For each referral in the list to Every referral done (sequence)
  • Create the chart to Stop the run (exception: not confirmed)

The map follows the emitted edge targets. It keeps each loop and exception route visible.

Test copy of OpenEMR, synthetic patients. Recorded values are left out. This replay combines four separate runs; the loop drawn around them was not itself run or qualified.

It reads the saved record back

Claude Sonnet 5.5 entered referrals into a test copy of OpenEMR with computer use, while we made some saves fail in the database.

30 of 30
bad saves the agent reported as done
25 of 30
caught by reading the record back
0
false alarms on clean saves
Injected fault, 5 times eachThe agent saidThe read-back found
A field blanked
A value changed
A duplicate chart
The save never landed
Another chart deleted
Another chart edited
Clean saves

Reported as done Caught as a bad save Confirmed as saved correctly Not caught

The 5 it missed edited a different patient’s chart. A later version of the check caught all 5 on rebuilt runs. Claude Sonnet 5.5 computer use alone, synthetic patients, faults injected only in the test database, 19 Oct 2026. See the runs

$0 in model fees per run. The compiled program calls no model.

Watch it catch a bad save on your computer

In about a minute, it runs a clinic app with synthetic records on your computer, makes one entry that saves and one whose save fails on purpose, and shows OpenAdapt stopping. No account, and nothing is uploaded.

uvx --python 3.12 openadapt@latest quickstart --break-it

For teams building AI agents

Your agent supplies the values, and the program enters them in the app

Your agent calls a compiled program by name. The program enters the values, reads the record back and returns the outcome.

YouEnter today’s 25 referrals.

AI agent

Ready

0/25
Enter referralTaught · 41 steps
View call
enter_referral(
  fname: "Nadia",
  lname: "Brennan",
  DOB: "1971-04-19",
  …
)

OpenAdapt step 0 of 41

Screens sent to a model
0
Model cost
$0.00

The agent passes the fields and gets back a status. No screen goes to a model.

Sped up, synthetic patients. Agent alone is Claude Sonnet 5.5’s real runs on the same patients. Both show 7 real runs from 14 Oct, repeated to fill 25.

Start in a minute

One command for Claude Code, or a few lines of Python for your own agent.

Claude Code: one command

claude mcp add openadapt -- uvx --python 3.12 openadapt-agent serve

Adds OpenAdapt to Claude Code. With no workflows of your own, it serves a sandbox workflow.

Your own agent: a few lines of Python

from openadapt_flow.jit import compile_agent program = compile_agent(my_agent, task=task, app=app, verifier=verifier, policy=policy, model_key=key) program.run(note="Follow up in two weeks") # first call: your agent does the task once program.run(note="Recheck in ten days") # after a person approves: the program, no model calls

Your agent does the task once. After a person approves it, OpenAdapt runs it as a program with no model calls and reads each save back.

For regulated settings, such as healthcare and insurance

OpenAdapt runs on a computer you control

It signs in to your system from there, and your records stay there. It suits work where a wrong entry is costly, such as a clinic’s charts or an insurer’s claims.

A person decides each save
By default, each save waits for a person, with a picture of exactly what will be saved.
The record is read back
After a save, OpenAdapt checks the saved fields through a separate read-only login.
The run report is signed
Anyone can check its signature in a browser.

OpenAdapt Desktop's phone page, 18 Oct 2026. The link to the computer is simulated; the card shows a real frame from the referral run above.

Our test runs and results, with their evidence

Our test runs use test systems and synthetic records, and each is published with its evidence. Your own runs stay private.

Each run ends with a receipt you can check

A receipt is a signed, itemized record of one run: what it changed, how each change was checked, and who signed it. It doesn't grade the run.

Referral 1 · receiptSaved and checked4 runs3 saved and checked1 stopped before savingOpen the receipt and check it in your browser →
How a receipt works

When a run ends, OpenAdapt writes a report of what it did and signs it with a key on the computer that ran it. Each mark on the receipt is a step or check from that report, and a stopped run gets a receipt too.

If any byte of the report changes afterward, the signature check fails. The signature shows that the report hasn't changed since the run; it doesn't grade the run.

Test copy of OpenEMR, synthetic patients.

Sources

  1. OpenAdapt scoreboard, the agent's claims versus the read-back, run openemr_referral/2026-10-19-a (SCOREBOARD.md). Claude Sonnet 5.5 did the entry alone with computer use, on a local OpenEMR test copy with synthetic patients. We injected faults only in the test database, in these ways: a field blanked, a value changed, a duplicate chart, a save that never landed, another chart deleted, and another chart edited. Each fault had 5 items, and there were 5 clean items as well. The agent claimed success on all 30 bad saves. verify-run caught 25 of the 30. It confirmed 5 of 5 clean saves, with 0 false alarms. The 5 misses were edits to another chart, and a later version of the check caught those on rebuilt state (5 of 5). The runs weren't timed.
  2. OpenAdapt quickstart with --break-it (openadapt launcher 1.17.0), MockMed synthetic clinic app, 18 Oct 2026 (the run's reports): the normal entry ended Done and checked by a read of the saved note; with the save rejected on purpose, the screen looked the same as the good save, and the run stopped and reported Check the record without retrying.

Pick a data-entry task your team does by hand every day